Published online 20 May 1999 | Nature | doi:10.1038/news990520-7


Making sense of sense

The hippocampus is an area in the centre of the brain that is known to play a crucial role in human long-term memory. But memory is not a single mechanism: it is a product of a suite of subsystems, each powered by different neuronal networks. Precisely where the hippocampus fits in to this suite - and what it actually does there - is not so clear.

Now a new study by Katharina Henke and colleagues, of the University Hospital in Zürich, Switzerland, adds to a growing body of evidence that the hippocampus is heavily involved in assigning meaning to concepts or episodes to be learned or recalled, and establishing mental connections between them.

As the researchers explain in the Proceedings of the National Academy of Sciences (11 May 1999), they presented twelve healthy young men with a variety of verbal learning and memory tests while they scanned their brain activity by using a method called positron emission tomography (PET). PET scans are effectively a ‘window’ on local increases in blood flow within the brain: blood flow is a measure of activity.

The tests involved learning pairs of familiar or unfamiliar, semantically unrelated words. The learning process involved the creation and memorizing of associations between the words, according to one of three strategies. The so-called ‘Associative Word Learning’ test, for example, required subjects to decide whether or not two abstract nouns fitted together in any way. In another task entitled ‘Deep Single Word Encoding’, participants were asked to decide whether or not one or both of a pair of words were ‘pleasant’ or ‘unpleasant’. And, in the final test, ‘Shallow Single Word Encoding’, Henke’s group asked their subjects to count the vowels in a pair of words.

Three minutes after each exercise, the researchers presented their subjects with pairs of the words and asked them to decide whether they were seeing old or new combinations - a task that would reveal whether associations had been memorized.

The PET scans revealed that blood flow in the hippocampus increased only when the brain was filing or remembering words according to their semantic relatedness. Other variables such as the novelty of a word, its appearance (the vowel test) or the emotional reaction it elicited (the pleasant/unpleasant exercise) made no difference to hippocampal activity. These results corroborate a previous study by the same group on learning pictures, rather than words.

Typically the hippocampus has been implicated in the learning of complex, unfamiliar and non-verbal ideas. This could be because, as Henke’s team suggest, “the memorisation of such material strongly requires the binding of sub-components into a complete mental representation.”